creating a dummy variable

[Upananda Pani]

Dear All,

I want to create four new dummy variables from the variable REGION2 which will NORTHEAST (REGION2==1) MIDWEST(REGION2==2) SOUTH (REGION2=3) WEST(REGION2=4)

I would appreciate if you can help me with the command

With best regards,
Upananda Pani

[Hemanshu Kumar]

If you are doing this purely because you need the dummy variables for an estimation command (such as regress or logit), you do not actually need to create the dummy variables. You can use factor variable operators that will temporarily create the dummies needed for the command. See

help fvvarlist

so e.g. in your regression of say y on x and region, you can just write

regress y x i.REGION2

[Juta Kawalerowicz]

You can create dummy variables from your factor variable with

tab REGION2, gen(REGION2_dummy)

This way you get 4 new variables called REGION2_dummy1, REGION2_dummy2, REGION2_dummy3, REGION2_dummy4.

[Upananda Pani]

Hi Mr. Juta K.,

While I was trying to check your code, it is not working. Please see the following code I have used to create as per your suggesting.

tab REGION2, gen(REGION2_dummy1)
tab REGION2, gen(REGION2_dummy2)
tab REGION2, gen(REGION2_dummy3)
tab REGION2, gen(REGION2_dummy4)

rename REGION2_dummy1 northeast
rename REGION2_dummy2 midwest
rename REGION2_dummy3 south
rename REGION2_dummy4 west

[Upananda Pani]

I think this method is not proper as it is creating so many dummy variable not the way we want.

[Hemanshu Kumar]

Upananda Pani no, you just need to run the tab command once, and it will generate all four dummies for you. You are getting too many dummy variables (presumably, 16) because you are running the command four times. Just run the exact command that Juta gave you.